From: "Rich" 

This is a multi-part message in MIME format.

------=_NextPart_000_01D1_01C2875C.124A10A0
Content-Type: text/plain;
        charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

   I was referring to the Math.Round method.  Had someone other than me =
thought to look at the documentation it explicitly calls out the round = to
nearest behavior specified by IEEE 754.

   As for IEEE 754, I thought the issue is that an operation either =
complies or does not comply not a feature list that you implement or = not.

Rich

  "Randy H"  wrote in message =
news:3dcc73d5{at}w3.nls.net...
  AFAIK, No; but you made the statement that .NET follows the 754 spec, =
so it only goes to
  reason to ask how much of 754 did MSFT implement.  =20
    "Rich"  wrote in message news:3dcc6357$1{at}w3.nls.net...
       Is there a difference in round to nearest between the two?

    Rich

      "Randy H"  wrote in message =
news:3dcc5c7e{at}w3.nls.net...
      Which vintage of 754? 1985 or 1987?
        "Rich"  wrote in message news:3dcbfdfe$1{at}w3.nls.net...
           .NET follows IEEE 754 which specifies round to nearest as =
round to even.  3.45 rounds to 3.4.  3.55 rounds to 3.6.  You can look = up
IEEE 754 on the web if you are not familiar.

        Rich

          "chad"  wrote in message =
news:3dcbd8a3{at}w3.nls.net...
          i feel all dirty for making a post so pro microsoft like i did =
before (in
          the anti MS rant thread.)  they really are evidence of the =
adage: absolute
          power corrupts absolutely.

          so i just got this from one of the development vp's in my =
company, in an
          email entitled "only in Microsoft software."  i figured i =
would send it
          along, to help cleanse myself.

          .NET's Math.Round function rounds "down" 0.5 (i.e. 3.45 rounds =
to 3.4)
          SQL Server's decimal-rounding rounds "up" 0.5 (i.e. 3.45 =
rounds to 3.5)

          Another "win" for MS :(

          chad - feeling a little better now


------=_NextPart_000_01D1_01C2875C.124A10A0
Content-Type: text/html;
        charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable








   I was
referring to the =
Math.Round=20
method.  Had someone other than me thought to look at the =
documentation it=20
explicitly calls out the round to nearest behavior specified by IEEE=20
754.
 
   As for
IEEE 754, I thought =
the issue=20
is that an operation either complies or does not comply not a feature = list that=20
you implement or not.
 
Rich
 

  "Randy H" <randy_holcomb{at}attglobal.netmailto:randy_holcomb{at}attglobal.net">randy_holcomb{at}attglobal.net
A>>=20
  wrote in message news:3dcc73d5{at}w3.nls.net...
  AFAIK, No; but you made the
statement that=20
  .NET follows the 754 spec, so it only goes to
  reason to ask how much of 754 did MSFT =
implement.=20
    
  

    "Rich" <{at}> wrote in message news:3dcc6357$1{at}w3.nls.net...
       Is
there a difference =
in round to=20
    nearest between the two?
     
    Rich
     
    

      "Randy H" <randy_holcomb{at}attglobal.netmailto:randy_holcomb{at}attglobal.net">randy_holcomb{at}attglobal.net
A>>=20
      wrote in message news:3dcc5c7e{at}w3.nls.net...
      Which vintage of 754? 1985 or =
1987?
      

        "Rich" <{at}> wrote in message news:3dcbfdfe$1{at}w3.nls.net...
          
.NET follows IEEE =
754 which=20
        specifies round to nearest as round to even.  3.45 rounds =
to=20
        3.4.  3.55 rounds to 3.6.  You can look up IEEE 754 on =
the web=20
        if you are not familiar.
         
        Rich
         
        
          "chad" <chabonowaka{at}hotmail.com>=20">mailto:chabonowaka{at}hotmail.com">chabonowaka{at}hotmail.com>=20
          wrote in message news:3dcbd8a3{at}w3.nls.net...i
=

          feel all dirty for making a post so pro microsoft like i did =
before=20
          (inthe anti MS rant thread.)  they really are =
evidence of the=20
          adage: absolutepower corrupts
absolutely.so i just =
got=20
          this from one of the development vp's in my company, in =
anemail=20
          entitled "only in Microsoft software."  i
figured i would =
send=20
          italong, to help cleanse
myself..NET's Math.Round =
function=20
          rounds "down" 0.5 (i.e. 3.45 rounds to
3.4)SQL Server's=20
          decimal-rounding rounds "up" 0.5 (i.e. 3.45 rounds to=20
          3.5)Another "win" for MS
:(chad - feeling a =
little=20
          better=20
now

------=_NextPart_000_01D1_01C2875C.124A10A0--

--- BBBS/NT v4.01 Flag-4